1. Introduction

Roselle (Hibiscus sabdariffa L) family Malvaceae, known commonly as "Karkade”. It is known under different names in different countries viz roselle, razelle, sorrel, red sorrel, Jamaica sorrel, Indian sorrel, Guinea, sorrel, sour -sour, and Queens land jelly plant (Mahadevan et al., 2009; Morton, 1987). It is an important crop in tropical and sub-tropical regions. The economical part of the plant is the fleshy calyx (sepals) surrounding the fruit (capsules).In Sudan fully developed fleshy calyx is peeled off from the fruit by hand and dried naturally under shade to give the dry (calyx), which is the consumable product. The plant, normally grown as annual plant, is 0.5 to 2 meters in height. It has a bushy shape with some what dense canopy of dark green leaves. The colour of the calyx plays an important role in determining the quality of the crop. The crimson red colour is the characteristic and most popular and desirable colour of roselle while other shades and colors exist, including the white or greenish white colour. It is an important cash crop in Western Sudan, particularly in Northern Kordofan State where the largest area of roselle is grown, especially in Elrahad and Um- Rawaba areas. The crop is mostly produced in traditional growing conditions by small-farmers, depending on rainfall and natural soil fertility without using chemical fertilizers or insecticides (El Naim and Ahmed, 2010). Roselle has many industrial and domestic uses. Locally, in the Sudan it is used as a beverage, where the dried calyx is soaked in water to prepare a colorful cold drink. Traditionally the product has been used for medicinal purposes for relief of sour throat and for healing wounds as an anti-septic (Aziz, 2007). Mahadevan et al (2009) reported that, in many parts of the world leaves is consumed as green vegetable and the stem is used as a source of pulp for paper industry. Seeds used as a poultry feed and as an aphrodisiac coffee substitute (Anonymous, 1959, Khidir, 1997). The total cultivated roselle area in Sudan during the 99/2000 season was estimated as 140,000 ha (El Naim and Ahmed, 2010). The main production comes from Western Sudan States, and the most of the exported crop is grown in the Eastern Kordofan localities. Roselle is also scattered in the southern region and south Fung area and recently at Abu Naama in the rainfed central clay plains of Sudan (Mclean, 1973). Weeds have been defined as higher plants in the Agro-ecosystem which are not sown, undesired ‘’ out of place’ or generally as plants which do more harm than good, (El Naim and Ahmed, 2010). They lead to direct yield losses through competition with the crop for water, nutrients, light, space and/or carbon dioxide. This degree of damage is mainly a function of their number, biomass and leaf area –index as compared with that of the crop. Weeds have different competitive abilities, which determine their performance and potential of damage in given situations; most important are vigor, growth habit, seed production, regenerative capacities and time of germination. In addition to competition, weeds can interfere negatively with cultural and harvest practices and may be poisonous and /or harbor pests and diseases. On the other hand, .weeds may have positive properties, when stabilizing the soil adding humus and nutrients to the soil (mainly) and extracting otherwise unavailable nutrients. In addition they might be alternative hosts for predators and used as sources of food, fodder, Fuel, .building material and for other technical purposes.(Weiikersheim, 1991). Groundnut for example is a field crop, but if it germinates in the subsequent culture, it is regarded as undesired ordinary weed. The objectives of this study were: to investigate the effect weeding frequencies on yield and Harvest index of roselle grow in sandy soil of Kordofan, Sudan in rain-fed.

2. Materials and Methods

A field experiment was conducted during season 2007/ 2008 under rainfed conditions in two fields naturally weeds infested in North Kordofan State, Sudan. The fields were: Abu Haraz, and Khor taqqat. The experiment was laid out in a randomized complete block design (RCBD) with four replications. The plot size was 5×4 meters consisted of 7 rows with 5 m along. The weeding treatments consisted of four levels (no weeding, weeding once (at 2weeks), weeding twice (at 2 and 3 weeks) and weeding three times (at 2, 4 and 6 weeks after sowing) designated as W₀, W₁, W₂ and W₃ respectively. Two widely used cultivated varieties of roselle (Elrahad, and Elfashir) were used in the experiment, designated as V₁ and V₂, respectively. Sowing dates on 11^th of July. Seeds were sown on rows at spacing of 70 cm apart and 40 cm within row, five seeds were placed in each hole. The plants were thinned to two plants per hole, two weeks later. The weed species found at each site were recorded at 15 DAS and then continued as interval of 14 days. Weeds counts made by placing the quadrate (0.5m x 0.5m) at random locations in plots repeated four times in order to obtain a reasonably good estimate of small weeds. The relative weed densities were calculated.

A destructive sample of five plants was taken at random from the five inner rows of experimental plot at maturity to measure the following yield attributes.

- Number of calyces per plant

- Calyces yield per plant (g): The calyces of five plants were peeled off from the capsules by using simple hand tools. The calyces were dried under shade to constant weight, and then average calices yield per plant (g.) was determined.

- Final calyces yield (kg / ha). Calculated by using the following formula:

- Harvest index was determined by using the following formula:-

Data were analyzed statistically using analysis of variance according to Gomez and Gomez (1984) procedure for a randomized complete block design. The differences of means were identified by Duncan^’s Multiple Range Test (DMRT) at P ≥ 0.05

3. Results and Discussion

3.1. Weeds and Stand

The majority of weeds in the experimental sites were the broad leaves (dicotyledons), while grasses (monocotyledons) found in a lesser density (Table 1). The presence and absence of weed species during the growing season are presented in Table 2. The majority of weeds in the experimental sites were the broad leaves (dicotyledons), while grasses monocotyledons) found in a lesser density. The dominant weed flora infesting Roselle (Karkade) during growing season were Cenchrus biflorus L (Alhuskaneet), Zornia glochidiata L (Sheilini) and Trienemara pentanture L (Alraba). They had relative weeds density of 27%, 21% and 11% respectively. El Naim and Ahmed (2010) found that the Cenchrus biflorus L was the most dominant weed in fields of Kordofan.

Table 1. Weeds Classification and Their Relative Density of non Weeded Roselle (Karkade) During the Growing Season in The Experiment Site

3.2. Growth and Yield Attributes

Weeds decreased plant height (Table 2). The significant differences in plant height among treatments may be attributed to the competition of weeds for soil moisture, nutrients, light and carbon dioxide. Weeding facilitates plants to have more resources for growth, these results agreed with El Naim and Ahmed (2010) who showed that, increasing weeding times increased plant height, due to efficient weed control.

Generally Elfashir variety (V₂) had significantly greater plant height than (V₁). Significant Differences in plant height among varieties were reported by El Naim and El Naim (2010), Cheweya (1992) and sulaiman (2005). The significant differences among weeding treatments in leaf area index (LAI) were observed in this study. Increased weeding frequencies increased leaf area index. This was due to better control of weeds. The reduced competition and increased availability of resources like nutrients, soil moisture and light paved way for higher leaf area per plant (leaf area index). These results are conformity with the findings of Kumara et al (2007) and El Naim and Jabereldar (2010).

Table 2. Presence (+) and Absence (-) of Weed Species During The Growing Season of un Weeded Roselle Weed species Local name Days from sowing 15 30 45 60 75 90 105 Zornia glochidiata. Sheilini + + + + + - - Abutilon figarinum. Alniada + + + + + + - Solanum dobium. Aljubain + + + + + + - Sesamum alatum. Simsim - + + + + + + Ipomea kordofana. Eltabar - + + + + + - Trienemra pentanture. Alraba - + + + + + - Corchorus olitorius. Almlukhia - + + + + - - Ocimum basilicum. Elryhan - + + + + + + Ipomea sinensis. Elhantoot - + + + + + - Rullia patula. Tagtaga - - + + + + - Cenchrous biflours. Alhuskaneet + + + + + + - Echinocola colonum. Aldiffera - - + + + + +

Table 3. Effect of Weeds and Varieties on Plant height and Number of Calyces per plant of Roselle

The weeds and weeding treatments had significant differences in mean number of branches per plant (Table 4). This result may be attributed to vigorous plant with less competition for light, nutrients, and free space in weed free environment. Similar results were reported by El Naim and Ahmed (2010).

In the present study, increased levels of weeding increased the number of calyces per plant (Table 3). These results agree with Small (2006), Adjun (2003), Baylan et al (1983) and El Naim and Ahmed (2010).

They showed that weeding three times resulted in high crop vigor score in number of calyces per plant. Moreover Ibeawuchi et al. (2005) reported that component of yield; pod number per plant can severely reduced by weed competition. The conservation water by plants due to weeding frequency could explain this result and this facilitating the growth essentials for regulating growth. The non significant differences among cultivars in number of calyces per plant might be to genotypic factors. Sulaiman (2005) reported that a significant variation among roselle cultivars in number of calyces per plant.

Table 4. Effect of Weeds and Varieties on Number of Branches per plant of Roselle

Table 5. Effect of Weeds and Varieties on Calyx diameter and Calyces Yield (Kg/ha)

Increased weeding frequencies increased the calyces yield per unit area (Table 5). This is because the number of calyces per plant and calyces diameter improved with levels of weeding, which led to increased yield (g /plant) and finally increased yield per unit area. Khater and Ahmed (1992), El Naim and Ahmed and El Naim et al (2010) found that the competition of weeds for light, water and minerals reflected on metabolic process of the plant and finally caused reduction in yield.

In this study, increased frequency of weeding increased the number of calyces per plant (Table 3). These results agree with the results of Small (2006), Adjun (2003) and Baylan et al. (1983) and El Naim and Ahmed (2010). They observed that weeds free plant resulted in high crop vigor score in number of calyces per plant. Moreover El Naim and Ahmed (2010) reported that component of yield; pod number per plant can severely reduced by weed competition. The conservation water by plants due to weeding frequency could explain this result and this facilitating the growth essentials for regulating growth. The non significant differences among varieties in number of calyces per plant might be to genotypic factors. Sulaiman (2005) reported that a significant variation among roselle cultivars in number of calyces per plant. Significant differences in the calyx diameter were observed among the treatments (Table 4). Weeds significantly lower the calyx diameter. This result agree with findings of El Naim and Ahmed (2010)

Significant differences among varieties in calyx diameter were obtained; by Mahmoud et al. (1996) and Sulaiman (2005). They reported that land races of roselle in Sudan have considerable variation in size of calyx. Chowdhury et al. (1995) reported that, weed free regime gave the highest yield than no weeding regime. Elfashir (V₂) had greater yield per unit area than Elrahad (V₁). This is because V₂ cultivar had the highest calyx’s number and calyx diameter. Similar findings reported were by Sulaiman (2005). Non significant differences were reported among the treatment in harvest index (Table 6). This is because the weeding frequencies and variety had similar effect in the economical and biological yield. These results are similar findings obtained by El Naim and Ahmed (2010) and El Naim et al. (2010).

Table 6. Effect of weeds and Varieties on Harvest Index of Roselle

Increased weeding frequencies increased the calyces yield per unit area (Table 5). This is because the number of calyces per plant and calyces diameter improved with levels of weeding, which led to increased yield (g /plant) and finally increased yield per unit area. Khater and Ahmed (1992) found that the competition of weeds for light, water and minerals reflected on metabolic process of the plant and finally caused reduction in yield. Chowdhury et al. (1995) reported that, weed free regime gave the highest yield than no weeding regime. Elfashir (V₂) had greater yield per unit area than Elrahad (V₁). This is because V₂ cultivar had the highest calyx’s number and calyx diameter. Similar findings reported were by Sulaiman (2005). Non significant differences were reported among the treatment in harvest index (Table 6). This is because the weeding frequencies and variety had similar effect in the economical and biological yield. These results are different from that obtained by Alam et al.(1995) in cotton.

4. Conclusions

Based on the results of this study, the weeding three times at 15, 30 and 45 days after sowing are effective to control weeds and recommended to improved yield of roselle crop in sandy dunes of north Kordofan (Sudan) under rain-fed conditions in traditional agricultural sector.